BERMS Sites Revisited: Footprint Climatology and 3D-LiDAR Data

We present results from a footprint climatology for the three Boreal Ecosystem Research and Monitoring Sites (BERMS, Canadian Carbon Program) located in the southern boreal forest, Saskatchewan, Canada. The footprint climatology in combination with a unique three-dimensional dataset of vegetation structure and elevation surrounding the towers opens up new opportunities for interpreting and understanding ecosystem processes.

For the present study, we derived a footprint climatology for the Southern Old Aspen, Southern Old Black Spruce, and Southern Old Jack Pine sites for the years 2005 to 2008. For each 30-min averaged sample period, the footprint area has been calculated taking into account atmospheric turbulence, prevailing wind direction and site characteristics. The footprint climatology thus provides spatial information on the actual sources/sinks distribution surrounding the eddy covariance tower on very high temporal resolution. The footprint climatology for the growing seasons of 2005 to 2008 was linked with airborne LiDAR (Light Detection And Ranging) data collected in August 2008. LiDAR data yield site and forest stand characteristics such as tree height, canopy depth, canopy fractional cover, a biomass-index, and elevation at very high, three-dimensional spatial resolution (35 cm). The combination of footprints, variability of the above site characteristics, and measured carbon dioxide and water vapour tower fluxes were used to determine the sensitivity of the fluxes to modest site heterogeneity. The results of this study illustrate that, even though the BERMS sites can be considered homogeneous compared with many other flux tower sites, both vegetation structure and small changes in elevation may impact net fluxes. Depending on site characteristics and weather conditions, the impact may be almost as significant as meteorological driving mechanisms. These findings suggest that upscaling approaches may be improved by weighting of fluxes and thus by accounting for within-site variability.